Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Calculating capacitor ripple current in output filter of DCDC module?

Status
Not open for further replies.
T

treez

Guest
Hello,
We are doing an SMPS using the DCM3623T50M53C2T00 Vicor DCDC module.

We need to know what is the ripple current in the external output capacitors that we must fit.

Please either state it (ripple current), or comment on our method of derivation…..


PSU Spec:
……………………………………….
Vin is 33V
Vout is 48V
iout is 6A
Crossover frequency of DCM module is 20kHz.
The DCM module switches at ~750kHz
This DCM module comprises an internal output capacitor of 11uF with 0.222R of ESR.
There must be a minimum of 200uF of external output capacitance (max of 2000uF)
DCM module is a ZVS type
……………………………………………….

Method of derivation of ripple current in output capacitor bank………..

Page 14 of the DCM datasheet shows the output voltage ripple…this shows that its not a sinusoidal type converter like an LLC…..
…Therefore, it must be a Phase shift full bridge, since that is a ZVS topology.

As such, for our 6A of average output current, the worst case for ripple current would be a Boundary conduction mode triangle wave of 12A peak……….such a current waveform (as attached) has an RMS of 6.92A….this would give a output capacitor ripple current of 3.45A rms…would you agree?

(this is obviously the worst case, ie boundary conduction mode inductor current in the PSFB output inductor)

DCM3623T50M53C2T00 Vicor DCDC module datasheet
https://www.vicorpower.com/documents/datasheets/DCM3623x50M53C2yzz_ds.pdf
 

Attachments

  • Output current of DCM module.jpg
    Output current of DCM module.jpg
    90.8 KB · Views: 150
Last edited by a moderator:

Hi,

As such, for our 6A of average output current, the worst case for ripple current would be a Boundary conduction mode triangle wave of 12A peak……….such a current waveform (as attached) has an RMS of 6.92A….this would give a output capacitor ripple current of 3.45A rms…would you agree?
I agree.

Klaus
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
V ripple is given as 1Vpp ( at a min load ) this is 350mV rms, most of the current ripple generated by the converter is soaked up by the 11uF internal MLCC's

you don't know the operating freq so you can't calc the converter ripple in the external caps, judging by the external ESR they require - they want to see the volt ripple on the o/p for control purposes - thus any old cap with moderate ESR up to 2000uF will do ...
 

just a note, most electro's are self resonant from 50 - 100kHz, i.e. above this they are inductors, they provide bulk storage for low frequencies only, if the Vicor sw freq is 750kHz, then the ripple could be 2x this - no electro will filter that ...
 

Thanks, the Fsw is not stated , i agree, but i found the figure hidden on page 14 which shows the vout ripple scope shot...which looks to be about 750kHz.
Yes i agree the electro's will look like inductors to the switching frequency....the crossover frequency of the converter is 20khz....so the electro's will see that...or rather, the converter will see the electro's.
....Just trying to piece together the vicor jigsaw puzzle.

- - - Updated - - -

The thing is, our 1mF electros have a "tan d " of 0.08...meaning an esr of micro Ohms....as such we wish to add to this.
 

the resistance of the cap at 750kHz ( it will be higher than ESR min at resonance ) will add vectorially to the XL at 750kHz to give a Z some what higher than the ESR...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Hi,

The thing is, our 1mF electros have a "tan d " of 0.08...meaning an esr of micro Ohms....as such we wish to add to this.
"tan d of 0.08" is only half of the information you need. You additionally need the tan d test frequency to be able to calculate ESR.
You talk about microohms... but if the value is given for 120Hz (which is not unusual for a 1mF electrolytic capacitor), then we are in the 100mOhms region.

Klaus
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
thanks, i was wanting to do the calculation with the "tan d" figure for 20khz...but its not available......though i believe "tan d" will be less at 20khz than at 120hz......which means the esr at 20khz will be less than at 120hz..........20khz is the crossover frequency of the dcdc module, so thats where i want to do the calculation.
 

Hi,

"tan d" will be less at 20khz than at 120hz.
Although I think that tan d will be higher at 20kHz ... the ESR may be lower...

Usually if the capacitor is not specified for 20kHz or higher it is not suitable for this frequency.

Klaus
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
though i believe "tan d" will be less at 20khz than at 120hz
No, no. You can review typical tan d and ESR characteristic from several manufacturers, e.g. Epcos/TDK. Tan d is roughly proportional to frequency.

sually if the capacitor is not specified for 20kHz or higher it is not suitable for this frequency.
Won't say so. But it's standard type with higher ESR and lower ripple capability.
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Thanks, with tan d being proprotional to Freq , that makes the ESR readings a lot lower.......taking the vishay 038 series electros...
https://www.vishay.com/docs/28309/038rsu.pdf

...the 220uF , 100v version would have esr of roughly 36milliohms, and the 1mF, 100v version 8 milliohms (taking 'tan d' to be 13.333 at 20khz when 'tan d' was 0.08 at 120Hz.)

I must admit its still tight and id still like to "add ESR"...partly because of the wide bandwidth of these vicor DCM modules, (20khz crossover) and adding ESR to tame it will help i think.
 

just for clarity, Tan-delta is used to calc losses, capacitor VA x tan-delta = watts

i.e. 2Vrms x 2A rms = 4 VA in the cap x tan-delta of 0.1 = 400mW in the cap ...
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top